Week 2: Dealing With Data

Hi all,

Welcome back! In episode 2 of AP Research, I started sampling and working with my data. This week was mostly spent sitting in front of a database and coding each image. To walk you through, I decided to sample 30 images from each company, randomly generating a number to select the respective image on the company’s folder within the SRITA database. To create the most objective coding process possible, I followed a method I found in a paper by Bengtsson et al. in which coding was a step-by-step condensation of the sample’s meaning. Each message is condensed until it is in a representative category; this way the risk of researcher bias is limited.

For instance, I might find that one image (which I do not want to show for copyright reasons) contains a vape pen, glasses, and an open book. I would condense this from “vape pen on book with glasses” to “vaping is associated with intelligence” to “intelligent people vape”. Ultimately, I’m left with the message that the ad sends to people, which can be compared with the message from the company. I’m still trying to work out how to make this even less subjective, but I may have to accept that some bias may come into play.

Next week I’ll be continuing to code data and perhaps start analyzing for trends. See you next time!